Development

Involved partners: Lead Beneficiary: Biomerieux (bMx)

Other participants: UCL, LUM, CHUV, GEN, ICH, INSEAD, AMC

The main objective of this workpackage is to prepare and optimise high-potential diagnostic technology platforms for use in diagnostic procedures in sepsis and to evaluate new clinical trial designs. This can be divided in two separate projects:

  • To optimise innovative immune functional assays that would be suitable to monitor the immune status and the risk of nosocomial infections of patients in critical conditions
  • To evaluate these prototypes in a pilot clinical study
  • To evaluate the sequential clinical trial design in sepsis research

Below these projects are described in a bit more detail:

Defining the role of molecules involved in regulation of inflammatory responses

Introduction:
The ESA-ITN project will make it possible to develop and standardize immune functional assays that are better suited to the constraints of the clinical hospital setting and particularly those of critical care conditions.

The projects:
Project 3.1:
This project aims to optimize new immune functional assay techniques to better characterize monocyte and lymphocyte anergy observed in septic patients and conform to the constraints of the hospital setting.

Project 3.2:
Taking the clinical context of sepsis and ICU into account, the global objective of this project will be to provide a prototype system allowing the digitalization and standardization of ex vivo immune functional assays at single-cell level using droplet-based microfluidics for studying single cells.

To evaluate these prototypes in a pilot clinical study

Introduction:
Molecular signatures derived from blood leukocytes provide more information about the nature and severity of the inflammatory response and the magnitude of organ injury than traditional single protein or gene biomarkers. These signatures can be used for identifying patients at risk of dying and/or sepsis complications, thereby allowing individualised preventive or therapeutic interventions.

The projects:
Project 3.3:
The main objective of this project is to identify molecular signatures within blood leukocytes, with three aims:
(1) to quantitate sepsis induced immunosuppression,
(2) to identify patients at high risk of developing (hospital-acquired) sepsis after ICU admission
(3) to stratify the individual patient with respect to risk of dying.

To evaluate the sequential clinical trial design in sepsis research

Introduction:
The problem of clinical trial design in sepsis are diverse. For example, patients enter sequentially and may receive one of two treatments (control or novel therapy arm). The rewards might be quantified by clinical benefit minus some function of cost of therapy. From patient to patient, the rewards may differ due to normal random variation. Parameters of the mean patient benefit may be unknown and are inferred (e.g. the presence or absence of biomarkers). The “big data” movement in data analytics has brought covariates that depend on patient characteristics into the picture. Similarly, recent work in stochastic simulation optimisation to incorporate Gaussian process regression and Bayesian value of information approaches to sequential sampling has reduced the number of samples required to identify the best alternative quickly.

The projects:
Project 3.4:
This project will build on our prior work in sequential modelling to
(1) adapt the Bayesian value of information approach for selecting the best stochastic alternative to the context of sequential clinical trials
(2) to use the Gaussian process regression model to extend the aggregation concept of Rigollet and Zeevi (2010) to potentially improve the use of continuous-valued diagnostic data
(3) to design a clinical trial algorithm based on the above results to allow for a small set of covariates to be accounted for in a clinical trial design
(4) to use a simulation study to allow for the new sequential clinical trial design to be compared with existing trial designs.

Project 3.5:
This project will be a collaboration of network partners at INSEAD, AMC, CHUV and ICH to make computer code (Matlab or R) of the mathematical development of the previous project, as well as computer code representing more classical approaches to clinical trial design.